[BensonBear]

Originally Posted by aliali47

I have many gigabytes of PDF books and articles and I am basically stuck at a limit of 50 on the reader at a time. It is simply too hard to locate anything when there are more.

I have a similar quantity of documents of various types, mostly pdf. I won't have all these on the Kobo (especially since it cannot deal with pdf files acceptably at current), but I think all that would be needed to find files is a slightly better search. Instead of searching for one string only in the search for a document, it should search for say, up to three strings, each of which must be present at the beginning of a word. I have this set up on my pc and it is trivially easy to find most documents. Eg, search "ch co wo" will find David Chalmers's "Constructing the World" on my system.

It would be very easy to add this to Kobo to generalize its trivially simplistic search which could still work the same if given only one string.

My question is: does the collections / bookshelves work with PDF files or epub only? Of course, I will try it myself but just wondered if there were any major issues with this before taking the plunge.

I had no idea why you would think it would not work with pdfs, so I tried it. Well, indeed, it seems not to work with pdfs! (I guess you are more cynical than I am). At least, when I tried to add a pdf to a shelf that already had an epub in it, it would not accept it. It would accept other epubs. I think probably no pdfs are accepted even in an empty shelf but I am not going to try because it is very difficult to do. The only way (I can see) to add books is to scroll through all your books until you find the one you want. My first pdf is 50 pages into the list (the last 10 from the end). Takes minutes to add the book. You cannot add a book when you are in it, or have located it with search (as far as I can tell, hopefully I am wrong about that).

The shelves have been left unenabled for a reason, as someone said. They appear not ready for use other than in trivial cases. (Also perhaps the tweak doesn't enable all the functionality).

It would be easier to make a program on your pc to create the shelves after figuring out how they were represented in the sqlite database.

A suggestion, linked to organising: would it be possible to create a tweak that allows the Kobo to sort by filename instead of metadata? I name my pdfs relevant to content but the metadata is often unrelated. That would mean renaming every single file on Calibre, manually, and I'm not prepared to do that.

Don't you mean that would require changing all the metadata to incoporate the info in the file name? Why would you have to rename the files? It would not be hard to extract the information from the filenames and stuff it into the metadata automatically. Calibre's "epub-meta" can easily be used in a simple shell script.

Another suggestion: enable highlighting/note taking on pdfs.

Indeed, it is very bad that one cannot do this. I saw someone saying that you could, but I have not been able to find it either.

I think these are the features that would change the Kobo from a toy to a researcher and scholar's companion.

Come on, the Kobo is not a "toy"!! But its targeted market is not researchers and scholars. I think it is people reading mostly novels and popular nonfiction on the subway, while away from home camping or travelling. A scholar or researcher will be doing most things at his desk(s), or at least with a more powerful computer.

That being said, it would be nice to have an entire scholarly library on the Kobo. There is enough space. But a lot more would be needed than just what you have mentioned. Although it would not be that hard to develop it. The annotations, already stored in one database, would have to be searchable, and groupable into groups that go across books, and serve as a general hypertextual index into the collection. A researcher/scholar would have to link together sections from many books along with his comments on them. A simple version of this, really, would not be that hard. If they made the nickel source open, I could do a basic version in a month.

For now, though, at least, we need to be able to easily READ pdfs. It is not possible, is it? How do you read a pdf now, compared to an epub? I find it horrible: first you have to enlarge it to the max in landscape mode and then painfully pan it up and down on the one page, and finally pan over to hit the advance-page region. I hope you agree that the first thing they have to do is to fix this, and make it work in maximally panned landscape mode exactly like epubs with a single tap to advance in the document, no panning required ever on a typical document.